| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Technical Briefs |
1
Institute of Clinical Chemistry, University of Regensburg, Franz-Josef-Strauss-Allee 11, 93042 Regensburg, Germany
2
Institute of Clinical Chemistry, University of Magdeburg, Leipziger Strasse 44, 39120 Magdeburg, Germany
a address correspondence to this author at: Universitätsklinikum Regensburg, Institut für Klinische Chemie und Blutbank, Franz-Josef-Strauss-Allee 11, 93042 Regensburg, Germany; fax 49-941-944-6202, e-mail
mustafa.porsch-oezcueruemez{at}klinik.uni-regensburg.de
Apolipoproteins define the functional properties of lipoprotein particles. In addition to their stabilizing features, several apolipoproteins have ligand functions. Some are also responsible for the modulation of enzymes involved in the homeostasis of lipid metabolism. Thus, apolipoprotein concentrations may provide essential information about lipid metabolism and associated diseases. There is evidence, for example, that apolipoprotein concentrations provide clinically relevant information concerning risk for coronary heart disease (1).
The determination of apolipoproteins in distinct lipoprotein subfractions narrows the conclusions that can be drawn for risk assessment as determined by or the discriminative power of HDL- and LDL-cholesterol (2)(3). In research, the determination of apolipoproteins encompasses their measurement in cell culture supernatants (4), lipoprotein subfractions separated by ultracentrifugation (5), isotachophoresis (6), immunoaffinity chromatography (7), or size-exclusion chromatography (8). The measurement of apolipoproteins in samples derived from such procedures requires appropriate methods reliable at concentrations below the physiological range.
Five analytical techniques are commonly used to quantify apolipoproteins. No delipidation is necessary in any of the advanced assays.
The use of radioactive reagents in RIAs is problematic (9). Radial immunodiffusion (RID) is simple to perform but time-consuming. Difficulties can occur when analyzing lipemic sera if the diffusion of the lipoprotein particles is complicated by particle size (5). Electroimmunodiffusion using the Laurell-Rocket technique (9) requires less time than RID, but large amounts of antibodies are needed. ELISAs provide several advantages (10), including good precision and a sensitivity comparable to RIAs. ELISAs are useful in routine clinical determinations because of the availability of automated methods. Nephelometry and immunoturbidimetry provide additional advantages in apolipoprotein measurement (11)(12). Whereas ELISAs are highly sensitive, nephelometry and immunoturbidimetry are superior with respect to precision, time, and cost. Therefore, nephelometry and immunoturbidimetry seem to be the most suitable methods for routine analysis of apolipoproteins.
To improve the relatively low sensitivity compared with ELISA, we optimized and evaluated immunoturbidimetric applications for low concentrations of the most frequently measured apolipoproteins, i.e., apolipoprotein (Apo)A-I, ApoA-II, ApoB, ApoC-III, and ApoE, attaining 12.5- to 45-fold higher sensitivities for these assays.
Commercially available assays for ApoA-I and ApoB ("ApoA1, immunologischer Trübungstest" and "ApoB, immunologischer Trübungstest") were obtained from Rolf Greiner Biochemica and modified to measure physiological concentrations of ApoA-II, ApoC-III, and ApoE. Assays were performed on a Hitachi 911 automated analyzer (Boehringer Mannheim).
The application characteristics of assays for physiological
applications and the final low-range conditions are given in Table 1
. Initially, samples were incubated for 5 min with buffer 1
("Immunofluid"; Greiner Biochemica) containing different
concentrations of polyethylene glycol (PEG 10000) dissolved and
stabilized in 100 mmol/L Tris buffer (pH 7.5) as provided ready
to use by the manufacturers. Sample volumes were increased 10- to
12.5-fold in low-range applications. Commercially available human
polyclonal goat antibodies specific against ApoA-I, A-II, B, C-III, and
E (Rolf Greiner Biochemica) were used without additional dilution in
all assays. Antisera were added after the first incubation step. The
resulting absorbance was determined after an incubation
interval of another 5 min.
|
Calibrators and control sera were purchased from Behringwerke ["N-Apolipoprotein" and "Apolipoprotein Control Serum CHD (human)"]. Analytical values of ApoA-I and ApoB were based on IFCC reference preparations (13). For ApoE, reference values provided by the manufacturers for nephelometry were used. The ApoC-III concentration of the N-Apolipoprotein standard was repeatedly measured by RID, and the mean value was used for calibration. Calibrators for the low-range applications were within the experimentally determined linearity range (see below) starting with a dilution of 1:16 of the N-Apolipoprotein standard. In all assays, a six-point calibration was performed. Sample concentrations were estimated by the logit-log method.
Low-range turbidimetry was compared with ELISA on an ES700 automated
analyzer (Boehringer Mannheim). The preparation and performance
characteristics of the ELISAs are described below. Polyclonal goat
antibodies used for turbidimetry were purified by affinity
chromatography, using HDL- or LDL-Sepharose 4B
(Amersham-Pharmacia-Biotech). After elution with 3 mol/L sodium
thiosulfate, antibodies were dialyzed against 1 mol/L
phosphate-buffered saline and aliquoted. Subsequently, the portion used
as primary antibody was stored at -20 °C until used. A portion of
the affinity-purified antibodies was biotinylated, dialyzed extensively
to remove the excess unbound biotin, and used as secondary antibody.
The reactivities of the primary and secondary antibodies were confirmed
by clearly visible precipitation bands within 48 h after diffusion
of each antibody lot against human serum on Ouchterlony plates.
Uncoated test tubes ("Enzym-Test System, uncoated tubes, naturel";
Boehringer Mannheim) were coated overnight with the primary antibody
diluted 1:1000 with 66.7 mmol
KH2PO4 and 66.7 mmol
Na2HPO4 (pH 6.0) and were
subsequently blocked with 20 mL/L bovine serum albumin (BSA).
Secondary antibodies were then added to the plates; anti-ApoB
antibodies were diluted 1:4000 with 10 g/L BSA, and the remaining
antibodies were diluted 1:2000. After the reaction,
streptavidin-coupled horseradish peroxidase and
2,2'-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid) were added
to the plates, and the resulting product was detected at 422
nm. Calibrators and control sera used were identical to those
used for immunoturbidimetry. Control sera were diluted 1:7500 in 70
mL/L BSA. Samples used for method comparison were diluted 1:750 to
obtain absorbances within the linearity range of the assays. All
samples belonging to the same comparison series were measured in the
same batch. The CVs for all of the ELISAs were 
5%.
The imprecision (as CVs) of the low-range assays (Table 1
) was
determined according to NCCLS protocol EP5-T2 (14). Although
the CVs for ApoA-I, A-II and B assay were <5%, those for ApoC-III and
E were higher. Because CVs are particularly critical at low
concentrations, we determined the functional assay sensitivity
contiguously in these ranges. The interassay CVs calculated from 10
measurements on 10 consecutive days are depicted in Fig. 1A
. ApoA-I, A-II, and B exhibited between-assay CVs <5% over a
broad concentration range. The CVs increased at 25 mg/L (ApoA-I), 10
mg/L (ApoA-II), and 15 mg/L (ApoB). For ApoC-III and ApoE, the CV
patterns differed considerably. The between-assay CVs for ApoC-III
steadily increased as the concentration decreased over the
entire linearity range, reaching 10% at 1.4 mg/L and 20% at 0.3 mg/L.
The between-assay CVs for ApoE were robust down to 1 mg/L. Below this
threshold, a large increase in the CV was observed. CVs exceeding the
20% cutoff were calculated for ApoE concentrations <0.4 mg/L.
|
Because in some of the assays, repeatedly measured blanks did not
differ from zero, the lowest concentration different from blank values
(detection limit) was estimated by the calibration curve method
(15), using the above-described calibration sera. In all
calculations, an 
error of 0.05 was assumed. Calibration curves were
generated by single measurements of six calibration points. The
following equation was used, assuming = ß:
 |
,ß and ß 
The resulting detection limits for normal- and low-range
applications are given in Table 1
. Improvements were 12.5-,
45-, 15-, 32.5-, and 22.5-fold for ApoA-I, A-II, B, C-III, and E,
respectively.
The assays were linear up to 65 mg/L for Apo A-II and 8 mg/L for Apo
C-III and ApoE. For ApoA-I and ApoB, a high-dose hook effect
attributable to antigen excess was observed starting at 275 mg/L
(ApoA-I) and 150 mg/L (ApoB), respectively. However, because of the
wide overlap between the detection limit of all normal-range assays
(Table 1
) and the linearity range given for the low-range applications
(Fig. 1C
), these problems could be circumvented by the simultaneous use
of both applications for each marker, which is recommended to detect
unexpected outliers. Because high salt concentrations also hamper
antibody reactivity, we checked the recovery of all five
apolipoproteins in high ionic strength environments. Potassium bromide
solutions with densities up to 1.25 g/L are used to isolate
lipoprotein fractions by sequential ultracentrifugation. Within the
range 1.0–1.25 g/L, no decrease of recovery was observed for samples
at the lower and upper ends of the linearity ranges of the respective
applications (data not shown). At higher salt concentrations, samples
had to be desalted before analysis because the recovery progressively
decreased.
The comparison of low-range turbidimetric assays with ELISA methods are
shown in Fig. 1B
. Immunoturbidimetry slightly overestimated
apolipoprotein concentrations except for ApoB. The most critical
discrepancy again was observed for ApoE, which exhibited a regression
coefficient of 0.869. In contrast, immunoturbidimetry underestimated
ApoB by 10%. Correlation coefficients of all method comparisons
performed underlined the good agreement of both methods.
Taken together, our results indicate that for apolipoprotein concentrations below the physiological range, our immunoturbidimetric methods offers features that are comparable to other more time-consuming, costlier techniques such as ELISA. Taking into consideration that both the low- and normal-range applications can be set up easily and in parallel on most currently used random-access analyzers, the proposed applications might be a helpful alternative to previous methods, even when there is a demand for the precise determination of low apolipoprotein concentrations. We therefore conclude that the described methodology is appropriate for rapid determination of apolipoproteins in research and routine diagnostic determinations.
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
, ApoA-I; 
, ApoB; 
, ApoC-III; 
, ApoE; 
, ApoA-II. Range
for ApoA-II (0–80 mg/L) shown on right. Ranges for ApoA-I
and ApoB (0–300 mg/L) and ApoC-II and ApoE (0–10 mg/L) shown on
left.